WILLIAM W. FROST Jr. M.D.
Washington, Pennsylvania

 

A pleasant 83 year old white male was seen by myself in consultation initially in March of 1989, at the Washington Hospital in Washington, Pennsylvania, for symptomatic left sacroiliac pain. The patient was seen briefly, noted to have normal lumbopelvic alignment, and was injected with Depo-Medrol and marcaine, 0.25%. He subsequently did quite well and was discharged from the hospital with his symptoms resolved.

He was not seen again in the hospital until he was admitted by his family internist on 21.4.91 for a chief complaint of sudden onset of lower back, left hip and groin pain, incapacitating and non-ambulatory. Because of the past history of severe chronic obstructive pulmonary disease with emphysema, the patient had been on long-term steroid therapy. There was also a known history of chronic renal failure, congestive heart failure, and chronic lymphocytic leukaemia, all relatively inactive at the time of admission. Prior x-rays in 1989 had shown
L3-4 degenerative disc disease and also some osteoporosis secondary to chronic steroids. Technetium bone scan in 1989 showed sacroiliac joint asymmetry.

The patient was seen in consultation at the Washington Hospital on 23.4.91 by myself. The internist had ruled out radiculopathy and neuropathy clinically, but because of the history was concerned about avascular necrosis of the hip. On physical examination during
consultation, tumbar mobility was restricted, especially in flexion, with a chief complaint of left groin pain increased with lumbopelvic flexion from the standing position. However, good relaxation was present in the lumbar paraspinals with no evidence of spasm or tightness.
The left iliac crest was noted to be slightly lower than the right in the functional standing position and there was 3+ to 4+ tenderness over the entire length of the left sacroiliac joint. A supine bridging maneuvre, followed by the knee chest position, followed by the patient
attaining the long sit position produced a significant 2.5 cm left lower extremity functional shortening. The patient was intact neurologically, and my impression was symptomatic left anterior innominate, with a shorter left lower extremity and secondary left sacroilitis. However, it was felt that avascular necrosis of the left hip should be ruled out and an MRI
was suggested.

Standard films of the pelvis and left hip showed no fractures, dislocations or significant bony abnormalities. Lumbar spine films confirmed the previously noted hypertrophic degenerative changes, especially at L3-4. An MRI of the left hip, including coronal plane studies, revealed
no evidence for changes in the marrow to suggest avascular necrosis or other marrow infiltrative processes. A further complexity is that the patient had been seen in consultation by a urologist during this admission and during the March of 1989 admission. He had had
prior work-ups for cancer of the prostate but they were previously normal. Because of this, the urologist suggested another Technetium bone scan. This was done on 21.4.91 and was found not to be consistent with any spinal metastatic disease. Technical factors prevented
evaluation of the sacroiliacs with confidence; the left sacroiliac joint may have had increased activity. A standard A-Pradiograph in the standing position revealed a functional leg length discrepancy, with pelvic iliac crest imbalance. Accordingly, my recommendation to the
attending physician was that the patient follow-up in my clinic, after discharge from the hospital, if no other serious explanation could be found for the patient’s incapacitating back pain. While still in the hospital, prostate biopsy revealed well-differentiated adenocarcinoma,
with potential for slow growth, and as noted previously no sign of any metastatic spread to the back or pelvis, by bone scan. Urology was reconsulted during the present admission and orchidectomy was done for prostate cancer.

The patient was subsequently seen in my outpatient office on 8.5.91. This time he was accompanied by his son. As part of our initial screening work-up, back patients are subjected to isometric computer evaluation using 10 functional work positions (NIOSH), with the
Dynatron 2000 isometric testing apparatus. The patient was only able to complete 8 out of 10 tests and simply was unable to perform the pull-in and push-out work position.

The patient was then brought in the examining room and his history was subsequently reviewed by myself. It was at this point that I was astonished to hear the following story. The son told me that his father, my patient, had had only one prior episode of severe,
incapacitating low back pain, left hip pain, and groin pain in his life. That evidently occurred in 1964 and the patient was subjected to a number of medical procedures, including electromyography, to rule out herniation and radiculopatby.

These were all negative. The patient’s son was living in Philadelphia, Pennsylvania at the time, and he and his father were desperate. He placed his father into a van and drove him (in the supine position) to Philadelphia, approximately a 6 hour ride. The son had heard of a
‘back wizard’, an orthopaedic surgeon named Dr. Myers. He brought his father to Dr. Myers, who evaluated the patient and told him that “prolotherapy might help your father”. The son went on to tell me that “Dr. Myers said that he was taught by a Dr. Hackett”. Further “Dr.
Hackett was taught by a Dr. Ongley”. The patient was immediately placed under general anaesthesia the same day he was seen by Dr. Myers in 1964 and underwent manipulation followed by approximately “70 to 80 injections of ligamentous points” the son tells me. The
response to injection prolotherapy was dramatic, the son states. He states that his father “couldn’t walk, was doubled over, and was totally incapacitated”. Yet on the same day, after receiving the injection and manipulations as outlined above, the patient himself was able to drive back to Pittsburgh, in the upright position, after treatment. The son stated that he thought Dr. Myers had injected “glucose and procaine”. However, the patient, 83 years old but with excellent memory, recalled that “there were 4 different medications involved”. They
were also given medication labels and information packs and literature by Dr. Myers, 27 years ago, as patient education materials during the prolotherapy regime. Evidently over the next 6 to 9 years a total of approximately 300 injections were given and this patient’s pelvis
was stabilised successfully.

After obtaining the above history, I then undertook a physical examination similar to that performed in the hospital the week before. Upon placing the patient into the knee chest position in an attempt to initiate the long sit position and biomechanically evaluate for leg
length discrepancy, there was a loud click which was observed in the presence of the patient’s son. The patient was placed in a standing position and he noted significantly decreased pain. His pelvis was level and it was apparent that he had spontaneously at least
temporarily reduced with regard to his pelvic malalignment. He was brought back to the computer analysis station and was asked to attempt to the pull-in and push-out manoeuvre and performed said quite easily without pain, complaints, or discomfort. On physical
examination, both in my office and in the hospital, I bad also noted gynecomastia, a finding which the family physician had felt was consistent with chronic obstructive pulmonary disease, long-term.

Accordingly, my final diagnoses were chronic low back; with symptomatic left anterior innominate dysfunction and a left lower extremity leg length discrepancy with shortening of 2-1/2 centimetres, as well as a very unstable pelvis. It was also felt that the patient had a
well-documented history of ligamentous prolotherapy 27 years earlier by Dr. Myers in Philadelphia, Pennsylvania. He had also been seen by me two years prior to the current intervention for active sacroiliac pain and had an injection for said pain. He was status-post
recent (21.4.91 admission) orchidectomy due to the history of the cancer of the prostate but had no evidence of metastasis. He was noted to have long-term steroid consumption by history, and also had physical findings consistent with feminisation.

DISCUSSION

The above case is extremely instructive and interesting from a number of viewpoints. A history of intact lumbopelvic function for a period of 27 years with the exception of isolated sacroiliac pain two years ago followed by recent recurrence of incapacitating left lower back
pain, hip pain and groin pain was noted in this patient. Long-term benefits from injection prolotherapy would seem to be confirmed by this history.

Additional ligamentous weakening factors such as history of chronic oral steroids and the history of severe progressive feminisation, both associated with the chronic obstructive pulmonary disease, further testify to the apparent effectiveness of Ongley prolotherapy.

Clinical reduction of the objectively confirmed left anterior innominate with left lower extremity shortening followed by the ability to perform pull-in and push-out testing further confirmed temporary reinstatement of functional integrity. The pull-in and push-out
dysfunction, in the above clinical context, has been demonstrated to be associated with structural abnormality (Frost, American Back Society Fall Symposium Abnormal isometric torque values in computerised individual muscle group functional capacity testing for
cervical, mid-back and low back injury San Francisco, California 7 December 1991). In this particular case, the pull-in and push-out were subsequently functional after spontaneous reduction, thereby confirming the above prior sacroiliac dysfunction.

Based upon all of the above, the revival of Ongley injection prolotherapy (T. Dorman, Treatment for spinal pain arising in ligaments-using prolotherapy: a retrospective survey, San Luis Obispo, California 25 April 1990) is to be hailed as a potential long-term solution
to chronic low back pain secondary to pelvic dysfunctions.

Dr. Bernard’s findings (ABS Spring Symposium, Toronto 1991 ) are also confirmed by this case report. Specifically, Dr. Bernard stated “when two or more sacroiliac joint stress manoeuvres were positive, and other well-recognised causes of back pain are excluded, the
diagnosis can be made with certainty. When recognised, a favourable therapeutic response may be anticipated with joint mobilisation, manipulation, or injection”. This was a retrospective study based on 250 patients.

Finally, and most importantly, this patient’s case is a piece of living medical history. It ironically highlights the frustration of some back caregivers currently working to ‘reinvent the wheel’ as well as ‘find the Holy Grail’. This is especially so as Dr. Ongley may very well
have found it, over a third of a century or so ago! I would be interested in the readers’ response and comments.

Address for correspondence:

William W. Frost, Jr., M.D.

(724) 228-0224
FAX: (724) 228-0285
E-Mail:harden@nauticom.net

Originally published in the Journal of Orthopaedic Medicine Vol. 16 1994 No 3

INTRODUCTION

Prolotherapy, the technology for strengthening lax ligaments, has found increased acceptance in recent years. However, despite its greater use, the mechanism of action of prolotherapy is not well understood. In the past few years a number of advances have been made in the understanding of wound repair. This author believes the increased knowledge which has been made available in the field of wound healing has application to a more complete understanding of prolotherapy. There follows a general discussion of wound healing and a hypothesis which provides a basis for understanding prolotherapy.

During prolotherapy, proliferating agents are injected directly into stretched or torn ligaments, resulting over a few weeks’ time in the loss of pain in the affected area and return to normal function of the associated painful skeletal articulation 1,2. Following injection of the proliferant, the clinician observes an immediate localized inflammation which diminishes gradually over several days. Patients are cautioned against taking aspirin or other antiinflammatory agents to relieve the discomfort. Over a period of several weeks or months, the pain with which the patient presented recedes and the treating physician observes a hypertrophied ligament with improved function of the articulation 3,4. (There is some debate concerning the relationship between mechanical dysfunction and the associated pain of ligaments involved. It is not the purpose of this discussion to venture beyond the biochemical mechanism of prolotherapy ).

How does prolotherapy occur and what is the underlying bio-chemical mechanism? Why is an injured ligament painful if injected with proliferant but an uninjured ligament remains relatively pain-free when injected? Why do non-steroidal anti-inflammatory agents (NSAI),
relieve the discomfort so rapidly? Is it reasonable that NSAI should diminish the therapeutic result? What is the purpose of the various mixtures that are used in treatment? How do they achieve their results? In order to answer some of these questions one must first understand, in a general way, how the healing process occurs. (For a more comprehensive description of the inflammatory process and its relationship to wound healing, the reader may consult some references listed at the end of this article).

WOUND HEALING PROCESS

Figure 1 shows schematically what happens during the healing of an injury 5,6. Where an injury occurs, cells are sheared and broken open; their contents are spilled into the wound site. Cellular debris and humoral factors attract an initial influx of granulocytes to the wound
site. These cells ‘debride’ the area through a combination of phagocytosis secretion of lytic enzymes and release of chemical agents such as hypochlorite and peroxide. Attractants for granulocytes may be present in extravasated cellular fluids such as clotted blood or they may
be released by injured cells. Many factors associated with cellular injury have been shown to be attractive for granulocytes. They range from thrombolytic peptide fragments, by products of the clotting process, to small chemical messengers derived from cell membrane lipids.
Larger proteins found within many cells have been demonstrated to be strongly chemotactic for granulocytes. Whatever the initial chemotactic signal, granulocytes are among the first cells to arrive at an injury site.

Figure 1     Summary of Wound Repair

Inflammation: early and late inflammation lead to granulation tissue formation. Granulation tissue is rich in fibroblasts and mononuclear cells which result in healing, through a process of matrix deposition and remodeling. The new matrix is composed of collagen and other
substances which are accumulated during a period of months, giving added strength to the injury (Clark 1985)

In addition to secreting chemical debriding agents, granulocytes are known to release their own chemotactic factors which recruit different cells to the injury site. Cells which arrive soon after the granulocytes are monocytes and macrophages. These large phagocytic cells
augment the actions of granulocytes, further ridding the injury site of cellular debris and opportunistic pathogens.

The initial phase of inflammation or granulocytosis lasts about three days (which interestingly is approximately the life-time of a granulocyte). During this phase, the second, overlapping, late inflammatory phase begins, involving monocytes and macrophages. The
late inflammatory phase; wherein large concentrations of monocytes and macrophages are observed, lasts for a longer period, about ten days. In both the early and late stages of the inflammatory process, polypeptides are secreted which attract fibroblasts to the injury site.

Among these factors, chemotactic for fibroblasts, are a number of polypeptide growth factors which have been demonstrated to be synthesized by macrophages. Fibroblasts are recruited to the area by these polypeptide growth factors and are stimulated to produce new
intercellular matrix, including new collagen. It is this new collagen which primarily gives strength to the healing wound; de-novo collagen synthesis and deposition is necessary for lasting wound repair.

A soupy indistinct mixture containing granulocytes, macrophages and infiltrating fibroblasts is known as granulation tissue; it is characteristic of a healing wound in the late inflammatory phase. As granulocytes and macrophages phagocytize the injured tissue at the
wound site, they simultaneously secrete polypeptide growth factors which attract and activate fibroblasts. Fibroblasts, in turn, deposit new intercellular matrix. Within the healing wound, as time passes, the granulation tissue matures, staining less pink and more blue (with haematoxylin and eosin staining) as more collagen is accumulated.

Ultimately, the granulocytes die and are removed by the macrophages which then slowly leave the wound site, while the fibroblasts remain, secreting a matrix of collagen, peptidoglycan, hyaluronic acid and other substances which increase the integrity of the injury site. Research indicates that there is a continual deposition and dissolution of tissue within the injury site. This “remodeling” process results from the conflicting actions of macrophages and fibroblasts, one enzymatically dissolving intercellular matrix, the other depositing it. The period of matrix formation and tissue remodeling continues for many months after the initial injury. In fact, in animal models, a healed wound does not acquire its maximum strength until about one year after injury, implying that collagen continues to be deposited and reformed for a period of months. This remodeling period may be important for the improved cosmetic appearance of healing wounds with time, allowing the initially less organized collagen to be rearranged into more functional orderly arrays.

In summary, three critical phases occur during the wound healing process: an initial inflammatory reaction which attracts other important cells to the injury site; a secondary inflammatory response in which macrophages secrete humoral factors which attract fibroblasts; subsequently, an infiltration and activation of fibroblasts which lay down new collagen, giving strength to the injury site. This process is called the wound healing cascade and, in the most basic form described above, is characteristic of all healing wounds. It is very important to note that each phase is dependent upon the previous phase for its initiation: tissue trauma leads to an initial influx of granulocytes; a large concentration of granulocytes attracts monocytes and macrophages; macrophages secrete polypeptide growth factors which recruit and activate fibroblasts; fibroblasts deposit the structural materials which repair the wound.

COLLAGEN BIOCHEMISTRY

What happens to the collagen which is deposited at the injury site? Very simply, new collagen shrinks. Collagen (see Figure 2) is synthesized as procollagen, a linear protein with large globular domains at each end of the linear portion 6,7. (Collagen biochemistry is a
complex subject, this discussion is simplified in the interest of clarity). The globular protein domains allow procollagen to remain soluble in cellular fluids. The linear domain is thereby prevented from associating with other procollagen molecules and precipitating from solution
before it is exported from the cell.

Once secreted by a fibroblast, however, proteolytic enzymes remove the globular domains, freeing each linear procollagen domain to combine with other linear procollagen domains or, perhaps, other components of the extracellular matrix. Some types of collagen form loose
web-like associations with other components of the extracellular matrix. Other types of collagen form more ordered macromolecular associations called tropocollagen; by winding around each other, like strands of rope, they form a triple helical strand of collagen (actually
tropocollagen is formed in the golgi prior to export from the cell, with larger scale ordering occurring outside the cell).

The association of three linear procollagen domains to form the tropocollagen triple helix occurs through hydrogen bonding. As one procollagen strand binds to two other procollagen protein strands, forming the characteristic collagen triple-helix, the water molecules which previously solubilized the polypeptide strands are replaced by protein-protein hydrogen bonds; water is excluded from the spaces between the linear polypeptide domains. At this point, the collagen triple helices, now outside the cell, combine with others in the
extracellular matrix and by a form macromolecular fibrils. As time passes, more water is excluded by these string protein-protein interactions. During this dehydration process, the protein becomes more compact. In addition, the fibrils become chemically crosslinked,
permanently bonding to one another. As the fibrils crosslink, they become hard almost crystalline. They lose even more water. They become the white acellular semicrystalline material which we observe in ligaments and tendons. The exclusion of water by proteinprotein hydrogen bonding brings about a collapse of the hydrated structure of the polypeptide which was previously solubilized; the protein occupies less space, it becomes smaller. If the new collagen is in a ligament, the ligament shrinks; it shortens as the intermolecular spaces collapse upon the spaces previously occupied by solvent.

Anyone who has seen the skin contract around a healing skin wound has observed that new collagen contracts during the late stages of healing. (Some other processes are active in the dermal intercellular matrix, including the production of myofibrils by myofibroblasts which
assist wound contraction; these processes may occur within ligaments also). Contraction occurs at any site where new collagen is produced, even within ligaments. This is the main point to remember in our discussion of collagen.

Figure 2

Collagen biosynthesis
Collagen in the extracellular matrix forms strong compact bundles of fibrils. These fibrils are constructed from tropocollagen units formed in the Golgi from the association of three procollagen polypeptide strands. (Carnell, Lodish and Baltimore 1986)

PROLIFERANTS

Now we can discuss the proliferants. Most basically, proliferants used for prolotherapy are substances injected into a ligament which lead to new collagen formation. The way that they do this is by initiating the first step in the wound healing cascade discussed: they cause local inflammation. Once the inflammation has begun and granulocytes and macrophages have been attracted to the injection site, fibroblasts will surely follow and deposit new collagen. The new collagen that is produced at the injection site undergoes contraction and pulls the ligament tighter. Proliferant solutions vary in the mechanism by which they cause localized inflammation but, in general, they all act by causing localized tissue trauma or irritation which initiates an influx of inflammatory cells. The exception to this rule is sodium morrhuate which may act as a chemotactic factor by a more direct mechanism. Perhaps, even more simply, proliferants are “inflammatory agents” which, by initiating the first step in the wound healing cascade, lead to fibroplasia.

IRRITANTS

The first class of proliferant solutions, called irritants or haptens, are exemplified by phenol, guaiacol and tannic acid. These substances are found for example in P2G (P25G) and plasma-QU. Each of these compounds has phenolic hydroxyl groups which are readily
oxidized to produce reactive quinone-like compounds (see Figure 3) which are known to directly alkylate the proteins on the surfaces of cells. By attaching themselves or their quinonoid oxidation products to the surfaces of cells at the injection site they either damage
the cells directly or render them antigenic. In either case, granulocytes and macrophages are attracted to the injection site and early inflammation occurs; in other words, the wound healing cascade is initiated.

Figure 3

Irritants
Agents with phenolic hydroxyl groups oxidize spontaneously to reactive quinoid compounds capable of alkylating cells.

Particulates

Particulates such as pumice flour are irritants of a different type. Small particles, on the order of 1 micron (about the same size as bacteria), are noted for their ability to attract macrophages. The macrophages phagocytize small particles in the same way in which they
ingest bacteria and cellular debris. This is the principle upon which cell sorters are used to separate macrophages from other cells; the macrophages are allowed to ingest magnetic particles and are then separated from other cells by a magnetic field. Once the macrophages
arrive at the injection site, ingesting the pumice granules, they likely secrete polypeptide growth factors which result in fibroplasia and new collagen deposition. Granulocytes also may be attracted to the cellular trauma caused by injection of particulates.

OSMOTICS

A second class of proliferants is characterized by osmotic shock. These agents act by dehydrating cells at the injection site. In osmosis, concentrated solutions cause a net flow of solvent across a semipermeable membrane from solutions which are less concentrated. In our case, the less concentrated solution is that found within living cells and the semipermeable membrane is the cell membrane; the more concentrated solution, the osmotic proliferant, causes a net flow of water into the injection site by removing water from living cells. Upsetting the delicate balance within these living cells causes severe, but localized, tissue trauma, No doubt many of the cells at the injection site are killed. Cells at the injection site, which are either morbid or dead, release cellular fragments (proteins, membrane
fragments and the like) which are attractive for granulocytes and macrophages. Thus, local tissue damage causes an influx of inflammatory cells and initiates the wound healing cascade. Osmotic proliferants, concentrated solutions of simple water soluble compounds, include concentrated glucose, glycerin or zinc sulfate. (Zinc sulfate may have other activities such as metallation of certain cellular proteins, causing secondary cellular injury; zinc binds very strongly to certain functional groups on proteins, particularly the sulfhydryl groups of
cysteine. Glucose may have secondary effects as well. For example, it is known that in diabetics high local glucose concentrations lead to non-specific glycosylation of some cellular proteins).

Figure 4     Arachidonic acid and its metabolites

Metabolites of arachidonic acid, a normal component of the cell membrane, are important mediators of the inflammatory process (Hood, Weissman, Wood and Wilson 1984)

CHEMOTACTICS

A third class of proliferants only has one member currently; sodium morrhuate contains the biosynthetic precursor to certain chemotactic agents which attract inflammatory cells 8,9. Sodium morrhuate is the sodium salt of the fatty acid component derived from cod liver oil.

Cold water fish oils are rich in polyunsaturated fatty acids such as arachidonic acid and related 20 carbon polyunsaturated fatty acids 10. (see Figure 4). These compounds are direct biosynthetic precursors to the mediators of inflammation such as prostaglandins, leukotrienes and thromboxanes 11. Cells use these compounds to communicate with other cells. For example, tissue trauma is thought to lead to the production of prostaglandins which promote vasoconstriction and swelling within the local area as well as attract granulocytes. It is thought that both the identity of the compounds produced and their concentration are important for intercellular communication. The documented powerful proliferant action of sodium morrhuate may be due to its arachidonic acid component being directly converted into prostaglandins and related mediators of inflammation. The observation that aspirin or ibuprofen (powerful inhibitors of key enzymes in prostaglandin biosynthesis) immediately eliminates the discomfort of sodium morrhuate injection corroborates this hypothesis.
Sylnasol may act similarly: the polyunsaturated fatty acid component being contributed by a plant extract rather than coming from marine oils.

Example: 20% Glucose Solution

As an example of one of these solutions, let us see what happens in the course of treatment with 20% glucose. Upon introduction at the injection site, the high osmolarity of 20% glucose immediately causes a new flow of water out of cells across their cell membranes into
the immediate vicinity of the injection site. The cells at the injection site become as flaccid and dehydrated as if they had been exposed to the air. These compromised cells become morbid and may eventually die. When cells are osmotically shocked, or upon cell death, the
integrity of the cell membrane is lost and factors are released which attract granulocytes, including fragments of cellular proteins and other cellular debris. Those cells not killed immediately release prostaglandins, leukotrienes and thromboxanes, the effects of which are
to attract inflammatory cells and to isolate the injury site via vasoconstriction. These small molecular weight mediators of inflammation are derived from the cell membrane fatty acid, arachidonic acid.

Arachidonic acid and closely related 20 carbon polyunsaturated carboxylic acids are transformed by an enzymatic process into a number of inflammatory compounds, including the prostaglandins, leukotrienes and thromboxanes. These biosynthetic derivatives of arachidonic acid are important chemical messengers used by cells for communication.
Certain enzymes in the biosynthetic process can be inhibited selectively, eliminating the production of many of these inflammatory mediators. For example, aspirin, acetyl salicylic acid, inhibits the enzyme cyclooxygenase which carries out a key biosynthetic step in the
prostaglandin synthetic pathway. This is the origin of aspirin’s ability as a pain reliever.
Inflammation is painful; eliminate the inflammation and the pain is relieved.

If inflammation is the first step in the healing response, then anything which interferes with the inflammatory response inhibits healing. Many practitioners of prolotherapy have observed that aspirin or ibuprofen immediately eliminates the discomfort associated with prolotherapy but that the clinical result is correspondingly diminished. The ability of nonsteroidal anti-inflammatory agents to inhibit the healing response has been well-documented in experimental situations and has been observed clinically 12.

To continue, the first biochemical step generated by using 20% glucose is localized tissue trauma, based upon osmotic shock, which initiates an inflammatory reaction, the first step in the wound healing cascade. Granulocytes, attracted to the injection site by cellular debris and chemotactic agents like prostaglandins, begin to chemically ‘debride’ the injection site and, during the process, secrete humoral factors chemotactic for macrophages. The macrophages phagocytize cellular debris and dying granulocytes, simultaneously secreting polypeptide growth factors which recruit and activate fibroblasts. The fibroblasts infiltrate the injection site and begin to secrete new collagen which, over time, forms a stronger, tighter, thicker ligament, through a process of crosslinking and dehydration. Photomicrographs of histologically prepared samples taken from injection sites confirm that granulation occurs at the prolotherapy injection site 1,2,3,4,8,9.

An interesting secondary effect of concentrated glucose solutions stems from the known ability of glucose, at higher than normal concentrations (for example in diabetics), to glycosylate tissues and make them appear foreign to the immune system. This non-specific
process, if it occurs at the prolotherapy injection site, would be expected to be immunogenic, inducing a localized inflammatory reaction. such a mechanism also may be operating when 20% glucose is used as a proliferant. Whatever the initiating mechanism, the subsequent
inflammatory reaction and the consequent wound healing cascade lead to fibroplasia in due time.

A third possible explanation for the action of concentrated glucose may be that it provides a gradient of a desirable nutrient which attracts mobile cells of the body’s repair apparatus to the injection site. Most cells use glucose for energy and cells of the immune system may
well be attracted to this locally concentrated energy supply.

SUMMARY AND CONCLUSIONS

In summary then, the basic mechanism of prolotherapy is not so difficult to understand, once the wound healing process is understood. Proliferants are injected which initiate local inflammation. The inflammation launches a wound healing cascade resulting in the
deposition of new collagen and a hypertrophied ligament. New collagen loses volume and contracts as it matures. The hypertrophied ligament is not only more robust but also, paradoxically, tighter due to the contraction which occurs with recently deposited collagen.

Why was the proliferant treatment needed if inflammation (following the initial injury) is naturally followed by repair? Perhaps the repair was incomplete following the initial injury or perhaps modern medical treatment interfered with natural healing. Many clinicians advise
their patients presenting with recent trauma, back pain or other joint pain to take antiinflammatory drugs. In light of what we know about wound healing in general, it may not be advisable to interfere with local inflammation immediately following a ligament injury. With
inflammation, the necessary sequelae which culminate in healing may not be initiated.

We now know that agents which inhibit inflammation also inhibit the wound healing cascade. Recent research has detailed the inhibiting effects of aspirin and related drugs upon healing. To attain a thorough initial inflammatory reaction, patients might profitably avoid commonly prescribed anti-inflammatory agents: steroids, aspirin or other non-steroidal antiinflammatory agents. A robust initial inflammatory phase leads to a corresponding infiltration of fibroblasts, the cells which produce collagen, the major strengthening agent in a healed injury.

Non-steroidal anti-inflammatory agents are most often prescribed because they are though to be a safe and conservative treatment modality. However, research has shown that nonsteroidals are not without side effects. In addition to their well documented adverse effects upon healing in the alimentary tract, they may directly inhibit healing of injured ligaments. It is well-accepted that prostaglandins, the target of non-steroidal anti-agents, are mediators of many important physiological processes in addition to their roles in the wound healing cascade. They are, quite simply, part of the chemical language used for intercellular communication. Interference in such important physiological processes (for example, the control of smooth muscle contraction) may be without consequences. Though they are not now completely understood, the widespread function of prostaglandins and related metabolites of arachidonic acid cautions against indiscriminate systemic inhibition of their biosynthesis.

The associated pain of the initial injury or the related discomfort of prolotherapy, which in an artificial way reintroduces the inflammation of the initial injury, is an important signal that healing is underway. Analgesics are available which reduce pain without direct interference in the inflammatory process. Such analgesics might be better employed in situations where a wound healing response is desired.

At some point in the future, chemotactic factors and polypeptide growth factors, produced by genetic engineering technology, will become available to the clinician 5,13. When this occurs, clinicians may be able to recruit fibroblasts directly to an injured ligament and perhaps avoid the discomfort of an inflammatory response. Until these days arrive, the astute clinician can stoically recite the old dictum that nothing worth achieving is obtained without sacrifice. Or in other words, proper healing does not occur without a certain amount of inflammation and discomfort.

REFERENCES

1 GS Hackett Ligament and Tendon Relaxation (Skeletal Disability) Treated by Prolotherapy (Fibro-Osseous Proliferation) Charles C Thomas Publishers Springfield, Illinois 3rd Edition 1958

2 GS Hackett and DG Henderson Joint stabilization: an experimental, histologic study with comments on the clinical application in ligament proliferation Am J of Surg 80 968- 973 (1955)

3 MJ Ongley, RG Klein, TA Dorman, BC Eek and LJ Hubert A new approach to the treatment of chronic low back pain The Lancet July 18 143-146 (1987)

4 RC Klein, TA Dorman and CE Johnson Proliferant injections for low back pain: histologic changes of injected ligaments and objective measurements of lumbar spine mobility before and after treatment J of Neurol and Ortho Med and Surg 10:2 123-126 (1989)

5 The Molecular and Cellular Biology of Wound Repair Eds:RAF Clark and PM Henson Plenum Press, New York (1988)

6 RAF Clark Cutaneous tissue repair: basic biologic considerations I J of Am Acad of Derm 13 701-725 ( 1985)

7 J Damell, H Lodish and D Battimore Molecular Cell Biology Scientific American Books New York (1986)

8 YK Liu, CM Tipton, RD Matthes, TG Bedford, JA Maynard and HC Walmer An in situ
study of the influence of a sclerosing solution in rabbit medial collateral ligaments and its junction strength Connective Tissue Research 11 95-102 (1983)

9 JA Maynard, VA Pedrini, A Pedrini-Mille, B Romanus and F Ohlerking Morphological and biochemical effects of sodium morrhuate on tendons J of Orth Research 3 236-248 ( 1985)

10 HN Brocklesby Structure of component acids and composition of marine anima! oils Section 2 in: Fisheries Board of Canada Bulletin No LIX The Chemistry and Technology of Marine Animal Oils with particular reference to those of Canada Ed:HN Brocklesby Ottawa (1941 )

11 LE Hood, IL Weissman, WB Wood and JH Wilson Immunology The Benjamin Cummings Publishing Company Menlo Park, California ( 1984)

12 JA Oates, GA Fitzgerald, RA Branch, EK Jackson, HR Knapp and LJ Roberts Clinical implications of prostaglandin and thromboxane A2 formation The New England J of Med 319:11 689·698 ( 1988)

13 JP Thompson, TR Oegema Jr and DS Bradford Stimulation of mature canine intervertebral disc by growth factors Spine 16 253-260 (1991 )

Originally published in the Journal of Orthopaedic Medicine Vol 13 1991 No 3

OBJECTIVES: To assess whether prolotherapy, an injection-based therapy,
improves elbow pain, grip strength, and extension strength in patients with lateral
epicondylosis.

SETTING: Outpatient Sport Medicine clinic.

STUDY DESIGN: Double-blind randomized controlled trial.

PARTICIPANTS: Twenty-four adults with at least 6 months of refractory lateral
epicondylosis.

INTERVENTION: Prolotherapy participants received injections of a solution made from 1 part 5% sodium morrhuate, 1.5 parts 50% dextrose, 0.5 parts 4% lidocaine, 0.5 parts 0.5% sensorcaine and 3.5 parts normal saline. Controls received injections of 0.9% saline. Three 0.5-mL injections were made at the supracondylar ridge, lateral epicondyle, and annular ligament at baseline and at 4 and 8 weeks.

OUTCOME MEASURES: The primary outcome was resting elbow pain (0 to 10 Likert scale). Secondary outcomes were extension and grip strength. Each was performed at baseline and at 8 and 16 weeks. One-year follow-up included pain assessment and effect of pain on activities of daily living. RESULTS:: The groups were similar at baseline. Compared to Controls, Prolotherapy subjects reported improved pain scores (4.5 +/- 1.7, 3.6 +/- 1.2, and 3.5 +/- 1.5 versus 5.1 +/- 0.8, 3.3 +/- 0.9, and 0.5 +/- 0.4 at baseline and at 8 and 16 weeks, respectively). At 16 weeks, these differences were significant compared to baseline scores within and among groups (P < 0.001). Prolotherapy subjects also reported improved extension strength compared to Controls (P < 0.01) and improved grip strength compared to baseline (P < 0.05). Clinical improvement in Prolotherapy group subjects was maintained at 52 weeks. There were no adverse events.

CONCLUSIONS: Prolotherapy with dextrose and sodium morrhuate was well tolerated, effectively decreased elbow pain, and improved strength testing in subjects with refractory lateral epicondylosis compared to Control group injections.